Research Paper Example
You should be aware by now that linear regression is a
very powerful statistical tool. However, one does not fully
see its power until one tries it out for oneself. Hence, the
last big computer assignment of the course will ask you to
find a problem which can be analyzed with regression,
collect data, enter the data into the computer and analyze
it with SPSS, and finally write a short paper describing
what you are doing and what your results mean. This paper
should be two to four typed pages.
Below are some examples you could do. However, I
encourage you to be creative. Find your own thing if you
- Sports problems: can one explain a pitcher's ERA on
the basis of hits, walks, and strikeouts he has per
inning? Can one explain walks on the basis of hits,
homeruns, and RBIs?
- Life expectancy: Countries around the world have very
different life expectancies. How does per capita and
population density affect it? Fertility rates are also
intriguing. Can they be explained with the same
- Politics: What factors determine whether a
congressional district votes Republican or Democratic?
Possible independent variables are percentage black,
percentage urban, average years of education, income, and
- P.E. ratios: What determines the price-earnings
ratios of stocks?
- College enrollments: How important is price in
explaining the number of students who enroll in various
colleges? What other factors matter?
- MPG that cars get: Weight should be important in
determining miles per gallon, but do engine size and type
of transmission matter too?
- Prices of anything: What determines the prices of
houses? of books? of cars?
For some problems you may have to collect the data
yourself. For others you may have to find it in a book. You
should try to get at least 30 observations.
The paper should explain what it is that you are doing,
where you got your numbers, what you expect to see in the
numbers and why, and finally an explanation of what the
statistical results tell you.
Once you get tuned in, there are many possibilities of
projects. The expected time to complete the project is three
to four hours.
(You must hand in your statistical output in addition to